openbuddy-yi1.5-34B-v21.3-32k Benchmark Results

openbuddy-yi1.5-34B-v21.3-32k — benchmark results

OpenBuddy's multilingual chatbot fine-tune of Yi-1.5-34B-32K with a 32k-token context window. Provider: Other. Released 2024-06-05. Access: Open.

Unified ELO 1500 ± 18, rank #800 of 1776 rated models, from 14 benchmark results.

Benchmark	Score	Metric	Percentile
Open Chinese LLM - C-Eval Semantic	89.73	Accuracy (%)	91.2
Open Chinese LLM Leaderboard	68.33	Average Score (%)	91.1
Open Chinese LLM - ARC Challenge	61.18	Accuracy (%)	90.1
Open Chinese LLM - GSM8K	64.29	Accuracy (%)	88.4
Open LLM Leaderboard - GPQA	13.2	Score	87.4
Open Chinese LLM - WinoGrande	68.19	Accuracy (%)	86.1
Open LLM Leaderboard - BBH	45.64	Score	85.8
Open LLM Leaderboard - MMLU-Pro	39.99	Score	85
Open Chinese LLM - HellaSwag	67.34	Accuracy (%)	83.1
Open Chinese LLM - CMMLU	69.05	Accuracy (%)	82.2
Open LLM Leaderboard - MuSR	14.69	Score	79.8
Open Chinese LLM - TruthfulQA MC	58.52	Accuracy (%)	77.4